Fermilab Computing Division

CS Document 3940-v1

24/7 operation of the Tier-0 processing system for the CMS experiment

Document #:
CS-doc-3940-v1
Document type:
Conference
Submitted by:
Oliver Gutsche
Updated by:
Oliver Gutsche
Document Created:
23 May 2010, 10:01
Contents Revised:
23 May 2010, 10:01
Metadata Revised:
23 May 2010, 10:01
Viewable by:
  • Public document
Modifiable by:

Quick Links:
Latest Version

Abstract:
The Tier-0 processing system represents the initial stage of the multi-tiered computing system of CMS. It takes care of the first processing steps of recorded proton-proton collisions at the LHC at CERN. The automated workflows running in the Tier-0 contain both low-latency processing chains for time-critical applications and bulk chains to archive the recorded data offsite the host laboratory. The Tier-0 system was successfully used during the cosmics data taking campaigns in 2008/2009 and the first proton-proton collision end of 2009. In 2010, The Tier-0 system was feature complete and processed all 7 TeV collision data taken with the CMS detector. This talk will summarize the operational experience gained during the 2010 data taking run. We will discuss lessons learned both for the usage of the infrastructure at CERN and the implementation and adaptation of the workflows to be run by the Tier-0 system. This talk will present a detailed analysis of latencies and efficiencies of the various processing workflows and will summarize the operational procedures put in place to guarantee a 24/7 availability of this initial processing step of the CMS data taking chain.
Files in Document:
Keywords:
CMS DataOperations
Associated with Events:
CHEP 2010 held from 18 Oct 2010 to 22 Oct 2010 in Taipei, Taiwan
DocDB Home ]  [ Search ] [ Authors ] [ Events ] [ Topics ]

DocDB Version 8.8.9, contact Document Database Administrators